UWB/GNSS/IMU Integration For Outdoor And Indoor Seamless Positioning

With the development of emerging industries such as artificial intelligence and autonomous driving,location-based services have undergone a rapid transformation from a single means to integrated Positioning,Navigation,and Timing(PNT)services.As the most common navigation and positioning technology,Global Navigation Satellite System(GNSS)can provide high-precision location services,but its signal propagation path is easily affected by occlusion and multipath errors in complex urban environments,making it difficult to provide reliable location services.Ultra Wide Band(UWB)positioning technology can provide centimeter-level indoor high-precision location services,which is an effective supplementary means in the environment of GNSS signal occlusion.However,as a radio positioning technology,UWB is also prone to system time latency and non line of sight(NLOS).In view of the above problems,in order to achieve continuous and reliable indoor and outdoor seamless navigation and positioning,this paper takes indoor high-precision UWB positioning technology as the research basis to carry out the theory and method research of UWB/GNSS/IMU integrated positioning,in order to realize high-availability location service in complex environments.The main research content and results of the paper are as follows:1.Aiming at the problem that the time latency of UWB sensors will greatly affect the positioning accuracy,the time latency of UWB sensors is firstly modeled,and on this basis,the UWB EKF algorithm is designed to take into account the time latency.In the static test,the position plane accuracy of the algorithm considering time latency is 0.073 m,which is 60.1%higher than that of the conventional UWB positioning method.According to the different error characteristics of abnormal observations,an improved UWB robust filtering positioning algorithm is designed based on IGGⅢUWB robust iterative filter.When the system contains general anomaly observations and system-level anomaly observations,the root mean square errors(RMSE)of the improved robust extended Kalman filter(EKF)positioning results are0.188 m and 0.270 m,which improve 20.3%and 45.6%respectively compared with the conventional robust EKF positioning.2.In complex environments,UWB sensors are susceptible to NLOS errors and signal loss.An improved UWB/PDR tight coupled positioning method is constructed,which integrates UWB ranging information and consumer-grade IMU,don’t solve UWB position separately,identifies UWB NLOS error based on theX~2 test,and uses Huber function to construct equivalent weight factor to adjust the measurement covariance matrix adaptively.The experimental results show that the improved UWB/PDR tight coupled positioning method not only suppresses the PDR drift error effectively,but also improves the positioning result accuracy by 47.7%and 24.7%respectively compared with the loose coupled under the conditions of UWB NLOS error and UWB signal loss.3.Construct a UWB/INS tight coupled positioning method fully considering the impact of time latency and NLOS error on positioning accuracy.Using TDOA observation of the time synchronization station to calibrate the time latency between base stations,the UWB tag equivalent time latency and the inertial navigation error state are solved together,the EKF is combined with the ranging innovation vector,and the multi-factor weight function is used to adjust the weight of UWB ranging information containing NLOS error adaptively.UWB/INS tight coupled positioning method considering the time latency and NLOS error has a plane accuracy of 0.124 m in the dynamic test of the NLOS environment,which improves 72.9%and38.9%respectively compared with the conventional UWB/INS tight coupled positioning method and UWB/INS tight coupled positioning method only considering the NLOS error.4.Considering the complementarity of different sensors and the demand for indoor and outdoor seamless positioning,a UWB/GNSS/INS integrated positioning method based on federated Kalman filtering is constructed,an indoor and outdoor seamless positioning experimental platform is built and corresponding experiments are carried out.The experimental results show that in the indoor and outdoor seamless areas,the average position errors in the E and N directions of the UWB/GNSS/INS integrated positioning mode based on federal filtering are 0.051 m and 0.039 m,which improve 19.4%and 39.1%respectively compared with the GNSS/INS tight coupled positioning.Due to the addition of UWB,positioning can be assisted when GNSS positioning accuracy is poor,so that the overall plane positioning accuracy of the system can be maintained within 10 cm.